monitoring driving behaviour in fuel consumption in light ... · anova1. vsp methodoloty when we...
TRANSCRIPT
![Page 1: Monitoring driving behaviour in fuel consumption in light ... · ANOVA1. VSP METHODOLOTY When we deal with real measurements on the road and taking into account the great variability](https://reader034.vdocuments.mx/reader034/viewer/2022052100/6039b51ba27cc523681068f0/html5/thumbnails/1.jpg)
1
Monitoring driving behaviour in fuel consumption in light duty diesel vehicles
Joana Peleja Madureira
Instituto Superior Técnico, Technical University of Lisbon, Av. Rovisco Pais, 1
Pav. Mecânica I, 2º andar, 1049-001 Lisboa – Portugal
Abstract
This work monitors and records the effects of a selected group of drivers and their driving behaviour in
the consumption of diesel fuel cars.
The challenge was based on the comparison of different driving behaviours and the individual analysis of
the drivers’ progress after they have adopted new techniques of driving.
The monitoring was accomplished by using the datalogger CarChip technology, manufactured by Davis
Instruments, and linked to the OBDII port of the vehicles concerned. This monitoring was carried out by
five drivers for a period of twenty days.
The initial phase comprises ten days which monitored the usual driving habits of the participants; the
following phase corresponds to the monitoring of the drivers’ behaviour after they have been fully
instructed about the eco-driving rules which aim for less fuel consumption and a reduction of the car
emission gases.
The evaluation criteria of driving behaviours, based on the eco-driving rules, encompass the driving
dynamics (acceleration, speed), idling time, correct use of the transmission and the analysis of fuel
consumption. Each car was checked to ensure that drivers would not be restricted by their cars
performance since these showed different engine power/weight ratio.
Two driving dynamic qualification and experience validation methodologies were developed using the
VSP (Vehicle Specific Power) and statistic methods, ANOVA1. The developed methods proved to be
efficient for it and it was possible to distinguish the driving behaviour of the different participants in these
two different phases.
The limitations of the vehicles demonstrated that the most aggressive driver seemed to be the only one
conditioned by the vehicle acceleration capacity. This one demanded more than 90% of the maximum
engine power in certain situations. Therefore one concludes that if he possessed a car with more engine
power/weight ratio, he would be even more aggressive.
It was also proved that the gentlest driver is gentle by choice, reaching the maximum of 60% of the
engine power of the vehicle. All participants showed alterations in their driving behaviour, becoming
gentler in accelerations and decelerations.
The unnecessary idling time was ignored by many drivers. In what concerns the use of fuel cut
mechanism situations it was verified that all drivers but one, benefited from this measure even before the
training lessons. The driver who didn’t benefit from this technology corrected himself in the second phase
of monitoring.
It was verified that all participants that showed significant alterations in their driving behaviour, becoming
gentler showed significantly consumption reduction to 29%.
Keywords: Monitoring, CarChip, Driving Behaviour, VSP, OBD
![Page 2: Monitoring driving behaviour in fuel consumption in light ... · ANOVA1. VSP METHODOLOTY When we deal with real measurements on the road and taking into account the great variability](https://reader034.vdocuments.mx/reader034/viewer/2022052100/6039b51ba27cc523681068f0/html5/thumbnails/2.jpg)
2
Objectives
This all project aims to monitor the driving behaviour effect on the consumption of diesel vehicles.
This work tries also to highlight the benefits, in consumption ratios, of the adoption of driving behaviour
practice, fairly simple to apply, and without involving costs in modifications either in technology or fuels.
This work intends to highlight the capacity of monitoring using small devices easily put to use in vehicles
– the CarChip technology. These devices enable the monitoring on the road and in real time, allowing us
to take into consideration all variables of the driving act. This work intends to contribute to the fields of
driving behaviour evaluation. Methodologies are presented to make possible the distinction of different
driving behaviours.
Monitoring
The monitoring applied in this project used the technology CarChip of Davis Instruments presented in
Figure 1.
The equipment is connected to OBII port of the vehicles. All vehicles manufactured from year 2000 are
equipped with this port. The available data from the OBII port vary from vehicle to vehicle.
The device can be programmed to register a maximum of five variables. Within the available parameters
one can find the car speed, the engine speed, the air flow rate, the air inlet temperature, the manifold air
pressure, the throttle position among other parameters.
It also allows the choice of time break to register variables. By default the speed variable is always
monitored second by second. The minimum time break to register the other variables is five seconds.
Figure 1. Monitoring device, CarChip (Davis Instruments)
Data Processing
Due to the complexity and huge amounts of vamples obtained it was necessary to use a software
programme that handled all the different observations and made a detailed data analysis. The chosen
software was MatLab
Firstly it was necessary to extract the data from the CarChip software, grouping all drivers’ journeys and
putting them into separated files.
Having the data processed and correctly read by MatLab it was necessary to separate them into two
types of files: one separating all journeys to enable analysis journey by journey; and another grouping all
the journeys together in order to accelerate the reading by the software when processing them all
together.
In short all the work processed in this project was through MatLab software. Several programmes were
elaborated to produce all the results.
![Page 3: Monitoring driving behaviour in fuel consumption in light ... · ANOVA1. VSP METHODOLOTY When we deal with real measurements on the road and taking into account the great variability](https://reader034.vdocuments.mx/reader034/viewer/2022052100/6039b51ba27cc523681068f0/html5/thumbnails/3.jpg)
3
Description of monitored vehicles
Five drivers were analyzed and their vehicles are shown in table 1.
Characteristics
Driver 1 Driver 2 Driver 3 Driver 4 Driver 5
Make Fiat Citroën Renault Peugeot Volvo
Model Croma C4 Picasso Clio 207 V50
Engine Diesel Diesel Diesel Diesel Diesel
Maximum Power [kW] 110(150 CV)
100(138 CV)
63 (85 CV)
50 (70 CV)
80 (109 CV)
Capacity [cm3] 1910 1997 1461 1398 1560
Transmission Manual Manual Manual Manual Manual
Driving Wheels Front Front Front Front Front
Acceleration
(0 to 100 km/h) (s)
9,6 12,4 12,7 15,1 12,1
Maximum Speed [km/h] 210 195 174 166 190
Combined Consumption [l/100km]
6,1
6.1
6,1 4,4 4,5 5,0
Urban Consumption [l/100km] 8,2 7,9 5,2 5,8 6,3
Extra Urban Consumption [l/100km]
4,9 5,1 4,0 3,8 4,3
Length [mm] 4783 4470 3986 4030 4522
Height [mm] 1603 1660 1496 1472 1457
Width [mm] 1775 1830 1707 1748 1770
Weight [kg] 1530 1656 1165 1160 1341
Engine Power / Weight [kW/kg] 0,072 0,060 0,054 0,043 0,060
Table 1 Description of monitored vehicles
Driving Evaluation Criteria
Tips on Eco-Driving Indicators
1- Driving in anticipation Reducing global acceleration
2- Accelerating and decelerating gently Reducing positive and negative acceleration
3- Avoiding idling situations Reducing situations in which vel=0 km/h
4- When descending and braking keep the car in gear
Reducing situations in which RPM=IdlingΛ vel>velmín km/h during ∆t > tmín seg
5- Driving in lower revolutions Reduction of high revolutions
6- Knowing how to analyze consumption Reducing high consumption
Table 2. Eco-driving criteria and respective numerical indicators
Table 2 includes the tips of a good eco-driver
The first two tips concern the driving dynamics, i.e. the gentleness/aggressiveness in accelerations and
decelerations. The third rule calls the attention to avoid unnecessary idling situations. The rules
corresponding to numbers 4 and 5 concern the correct use of the transmission. The reductions of
consumption (and the implied emissions) are the result of good driving practices – gentle dynamics and
the efficient use of the transmission.
The first eco-driving rule “driving in anticipation” does not differ much from the second rule “accelerating
and decelerating gently”. Basically both imply the same effect in driving – it is possible to state that by
carrying out the second rule you are in fact carrying out the first one. Gently accelerating and decelerating
one practices driving in anticipation, and therefore evaluated by the same mathematical indicator.
![Page 4: Monitoring driving behaviour in fuel consumption in light ... · ANOVA1. VSP METHODOLOTY When we deal with real measurements on the road and taking into account the great variability](https://reader034.vdocuments.mx/reader034/viewer/2022052100/6039b51ba27cc523681068f0/html5/thumbnails/4.jpg)
4
Therefore one is consequence of the other. However one must highlight that driving in anticipation makes
a driver more attentive to traffic conditions, avoiding more easily unpredictable situations and
unnecessary accelerations.
Firstly it was necessary to check if any of the drivers would be conditioned by the acceleration capacity of
the vehicle, jeopardizing the gentle qualifications and experience for the vehicle capacities. Bearing in
mind that the dynamics is obtained through the speed and accelerating variables and that the engine
power is obtained through the acceleration result for the speed and weight of the vehicle, it is possible to
check which engine power is demanded in each dynamic point of measurement (see equation 1). So a
methodology has been developed to evaluate how far the different capacities of acceleration of vehicles
would limit the drivers’ aggressiveness.
�� � ����� ��� � 1, … , � ��������� ��� �1�
Each value of Pi was divided by Pmax1 which achieved the demanded percentage to the engine in each
analyzed dynamic point.
Consequently there was a necessity of developing methodologies which evaluated the drivers’ dynamic
behaviour. Two methodologies of analyses: the VSP analysis and the one based on statistical methods,
ANOVA1.
VSP METHODOLOTY
When we deal with real measurements on the road and taking into account the great variability of driving
conditions, it gets complicated (contrary to the roll banks or in numerical simulation in which the variables
are more easily manipulated) to correlate specific events with emissions and consumptions and have
statistically sufficient data to validate and compare different performances of the vehicle and different
driving behaviours.
In each driving situation is concerned, be the vehicle going up a steep slope, doing a sudden acceleration
or far exceeding the speed limit, the power output of the engine can be the same in the different situations
as previously mentioned. The same will happen if it is in gentle mode, i.e. decelerating, or in a
descending situation without using the accelerator and in gear, i.e. in a fuel injection switch off situation.
The idea is to use a methodology that gathers demanding situations to the engines of equal power, i.e.
the same conditions of functioning, and to associate those to the qualifications and experience of the
driving dynamics. The VSP equation appears in the following form:
��� � ��� � ��sin �� � !".$%&�&'ê()�* + � !".*%$,-�(â.�), �/ �2�
Being:
VSP Vehicle Specific Power [W/kg]
v Vehicle Speed [m/s]
a Vehicle acceleration [m/s2]
φ Inclination of the road [radians]
Coefficient of resistance Coefficient of resistance to rolling
Aerodynamic coefficient
Since the present study is only for light vehicles the following values have been considered for the
coefficients of resistance to rolling and aerodynamics, 0.132 and 0.000302 respectively. It was not
possible to estimate the influence the slope had in this analysis for reasons of the limitations of monitoring
variables. One assumes flat topography or in other words that ascents were as many as descents.
The challenge in this moment is in using this VSP methodology to analyze the qualifications and
experience of the participants’ driving.
1 Pmax means the maximum engine power of each vehicle indicated by the manufacturers shown in table 1
![Page 5: Monitoring driving behaviour in fuel consumption in light ... · ANOVA1. VSP METHODOLOTY When we deal with real measurements on the road and taking into account the great variability](https://reader034.vdocuments.mx/reader034/viewer/2022052100/6039b51ba27cc523681068f0/html5/thumbnails/5.jpg)
5
When the VSP method was being analyzed, two influences divided into two components have been
identified: va (speed*acceleration) component and v (speed) component.
The va component has got a major influence for lower speed since the aerodynamic coefficient reduces
the speed weight greatly when this shows lower values. However the s component has got a higher
expression in high speed. Bearing in mind that the power of acceleration of the vehicles diminished at
high speed, the acceleration tends to be reduced in situations of excess speed.
One can make an analogy between the type of routes already achieved and expressed in the previous
paragraph: for typically urban routes the influence of the acceleration is more meaningful; for extra-urban
routes the speed component acquires a predominant importance.
Bearing in mind the objective of this project which was to monitor the influence of driving behaviour on
consumption, we used the VSP method to separate the influences of components va and s and attribute
a punctuation system to each form so that we could evaluate the different dynamic qualifications and
experience of the drivers.
Component va
It was initially necessary to visualize which was the influence of va in the expression (2) to evaluate the
influence of the component va in the driving dynamics. Therefore we sketched the isolines va. (see
figures 2, 3, 4, 5 and 6)
Va isolines have been sketched corresponding to the values -32, -16, -8, -4, -2, 2, 4, 8, 16 and 32 so that
to separate the areas in the dynamic qualifications and experience, giving each area a correspondent
punctuation shown in table 3.
va [0 2] [2 4] [4 8] [8 16] [16 32] [0 -2] [-2 -4] [-4 -8] [-8 -16] [-16 -32]
Score 100 50 0 -50 -100 100 70 40 10 -20
Table 3 Score table for the va intervals
Figure 2 va isolines combined with the qualification and experience dynamic for Driver 1
Figure 3 va isolines combined with the qualification and experience dynamic for Driver 2
![Page 6: Monitoring driving behaviour in fuel consumption in light ... · ANOVA1. VSP METHODOLOTY When we deal with real measurements on the road and taking into account the great variability](https://reader034.vdocuments.mx/reader034/viewer/2022052100/6039b51ba27cc523681068f0/html5/thumbnails/6.jpg)
6
Figure 4 va isolines combined with the qualification and experience dynamic for Driver 3
Figure 5 va isolines combined with the qualification and experience dynamic for Driver 4
Figure 6 va isolines combined with the qualification and experience dynamic for Driver 5
As one can see in table 3, the positive va scale is different from the negative va scale. This fact can be
justified for an aggressive acceleration can be more serious than a sudden deceleration in terms of
consumption effects. The braking in itself does not consume fuel although it causes rapid wear of the
brakes. The decelerations do not have a direct effect upon the consumption of fuel. However the
decelerations are serious because braking needs an acceleration to compensate the deceleration and
accelerations have a direct effect upon the consumption of fuel.
Component v
The component s corresponds to the influence of speed. The same procedure that has been done to the
component va, has also been attributed to a punctuation scale which benefits the vehicular traffic at low
speeds. The punctuation system works to guarantee the maximum punctuation of 100 points, penalizing
linearly from 90km/h, and once it reaches 120km/h the punctuation will attribute negative values
![Page 7: Monitoring driving behaviour in fuel consumption in light ... · ANOVA1. VSP METHODOLOTY When we deal with real measurements on the road and taking into account the great variability](https://reader034.vdocuments.mx/reader034/viewer/2022052100/6039b51ba27cc523681068f0/html5/thumbnails/7.jpg)
7
ANOVA1 METHODOLOGY
This work utilizes a statistical technique, the analysis of the variation (ANOVA) followed by a
complementary method known as Tukey-Kramer test. Through the statistical method of ANOVA1 one can
verify if, in fact, the drivers show significant behaviours different from one another which statistically can
be distinguished. The Tukey test appears as a complementary method to the ANOVA1, evaluating all
differences among the drivers with an interval of confidence of 95%.The ANOVA1 method was used to
analyze acceleration rate and negative acceleration rate and consumption rate separately.
The third rule “Avoiding idling situations” concerns those in which the engine is running but the vehicle is
stationary. So the idling time is an unnecessary waste of fuel. However all drivers that participated in this
project travel most of their time in urban areas where situations of traffic that imply using the idling occur,
such as obligatory stops at traffic lights. Obviously there are drivers who do not encounter traffic lights so
frequently.
It would not be a fair indicator since the idea is to evaluate the unnecessary time at the idling and not
penalize in time the obligatory tick over. With this type of monitoring it is not possible to investigate if we
deal with obligatory or unnecessary idling situations. If we had opted for estimating all moments in which
the engine is running and the vehicle speed is zero, it would have been an extremely subjective analysis
leading to fallacious results.
So we opted to evaluate the idling time at the beginning and end of each journey, situations in which none
of eco-driver wastes fuel.
The fourth rule “When descending and braking keep the car in gear” concerns the use of the fuel cut
mechanism. Nowadays all vehicles are equipped with this technology which makes possible in
descending and/or braking situations, keeping it in gear and taking the foot from the accelerator, leaving
in motion due to inertia results in zero consumption. So this rule allows the vehicle to save fuel in
situations where it is possible to profit from the fuel cut mechanism. The criterion found to evaluate this
situation resulted by contrast, i.e. penalizing situations in which the drivers, descending or braking, drive
in a non gear mode, i.e. gearstick in neutral. So it was necessary to identify numerical indicators to could
this behaviour as follows:
RPM= idling, ∆t>2sec e v>30km/h;
The first indicator is the one that characterizes the bad use of the transmission. The second one is used
not to encompass situations of changing gear. Sometimes when one changes gear the time taken by the
clutch is expressed in the engine speed a value equal to the idling time. For safety the time interval that
RPM is equal to the idling has to be superior to two seconds. The third indicator that imposes speed has
to be superior to 30km/h and is used not to confuse situations of non-use of fuel cut mechanism with
situations of pre-braking, since a traffic situation obliges the driver to immobilize the vehicle: this uses the
clutch and the engine speed descends to the idling engine speed.
The fifth parameter concerns the rule 5 “Driving in lower revolutions” was not evaluated in this project
because that criterion was not only related to the good use of the transmission but also with the speed
regime predominantly adopted. The drivers of this project did not travel the same routes as each other, so
some came out to be benefited. The drivers who drove with excess speed were doubly penalized, not
only by the RPM highest regimes but also by the VPS methodology for excess speed. Another reason, no
less important, is the fact that an eco RPM range is not universal. Because there is not a RPM eco
interval and because it is subjective to adopt one, since the vehicles, all diesels, are different and could
have different and efficient RPM ranges, we opted to abandon this criterion in the drivers’ evaluation.
The sixth advice of eco-driving, “Knowing how to analyze consumption” was embodied in this project to
reduce the natural consumption of fuels. This used as an indicator analyses if, actually, a reduction in the
consumption of fuels has been established when a more correct style of driving occurred. Basically this
indicator is the result of the application of all the other measures mentioned above. We felt the need of
![Page 8: Monitoring driving behaviour in fuel consumption in light ... · ANOVA1. VSP METHODOLOTY When we deal with real measurements on the road and taking into account the great variability](https://reader034.vdocuments.mx/reader034/viewer/2022052100/6039b51ba27cc523681068f0/html5/thumbnails/8.jpg)
8
comparing consumption values among the different participants. One should stress that all monitored
vehicles are European, manufactured after the year 2000. All participants declared both urban and fast
highways journeys. We used, mainly, the urban consumption used in combined journeys published by the
respective vehicle manufacturers, available in table 1, with the intention of normalizing the consumption
and comparing values.
Results
In this chapter the results for all the indicators previously explained are shown.
Figure 7 Ranking of the different drivers in both monitoring phases obtained for the positive component va through
the VSP method
In component va positive it was verified that all drivers improved in their driving behaviour. Driver 1
showed the highest improvement, 25,2%, followed by Driver 4 with 19,1%, then Driver 5 with 18,9%, then
Driver 3 with 11,0% and at last Driver 2 with 3,6%. Driver 2 improved, but it’s improvement was small
compared with the other four drivers.
Figure 8 Ranking of the different drivers in both monitoring phases obtained for the negative component va through
the VSP method
In negative component va the results showed an improvement for Driver 1, 31,6%, Driver 4, 21,9%. Driver
2 showed only 0,8% change in the second phase of monitoring. And for Driver 3 and Driver 5 the results
![Page 9: Monitoring driving behaviour in fuel consumption in light ... · ANOVA1. VSP METHODOLOTY When we deal with real measurements on the road and taking into account the great variability](https://reader034.vdocuments.mx/reader034/viewer/2022052100/6039b51ba27cc523681068f0/html5/thumbnails/9.jpg)
9
manifested a decrease of their driving behaviour concerning negative component va with 10,8% and
13,8% reduction, respectively.
Figure 9 Ranking of the different drivers in both monitoring phases obtained for the component v through the VSP
method
Analyzing the results for component v, it was verified that all drivers except Driver 1 reduced their speed
in the second phase of monitoring.
Figure 10 Results obtained for the positive acceleration variable (m/s2) through the ANOVA1 method
In the analysis of ANOVA1 results the scale is analyzed in the opposite way from the VSP methodology.
A reduction in acceleration rate is considered an evolution. So in this case, once again Driver 1 showed
the highest development with 19,5% reduction. Driver 5 and Driver 4 showed an improvement of 15,4%
and 11,6% respectively. Driver 3 also showed an improvement with 7,4% reduction. Again it was
observed that Driver 2 did not show much change in his/hers driving behaviour concerning acceleration
rate with only 2,7% reduction.
![Page 10: Monitoring driving behaviour in fuel consumption in light ... · ANOVA1. VSP METHODOLOTY When we deal with real measurements on the road and taking into account the great variability](https://reader034.vdocuments.mx/reader034/viewer/2022052100/6039b51ba27cc523681068f0/html5/thumbnails/10.jpg)
10
Figure 11 Results obtained for the negative acceleration variable (m/s2) through the ANOVA1 method
The analysis for ANOVA1 results for negative acceleration rate is similar to the ones obtained for positive
acceleration rate. Driver 1 showed a reduction of 17,1%, Driver 5 a reduction of 12,4%, Driver 3 9,6%
reduction, Driver 4 9,4% and Driver 2 with 4,3% reduction.
Figure 12 Average time spent in idling at the beginning and end of each journey from the different drivers in both
monitoring phases
For the average idling time (see Fig. 12), unnecessary at the beginning and at the end of each journey
one verifies that the drivers did not fulfill this eco-driving rule in the second phase of monitoring. In
general practically all with the exception of Driver 4 who reduced and Driver 5 who maintained, all
increase the idling time.
![Page 11: Monitoring driving behaviour in fuel consumption in light ... · ANOVA1. VSP METHODOLOTY When we deal with real measurements on the road and taking into account the great variability](https://reader034.vdocuments.mx/reader034/viewer/2022052100/6039b51ba27cc523681068f0/html5/thumbnails/11.jpg)
11
Figure 13 Percentage of distance run in out of gear mode from the different drivers in both monitoring phases
The use of fuel cut mechanism (see Fig. 13) one verified that the only driver that did not do it was Driver
4. Nevertheless Driver 4 showed a quite visible progress in the second phase of monitoring. In the first
phase he/she showed 21.88% of the distance run in neutral; in the second phase, after being enlightened
of the benefits of good use in fuel cut mechanism, he/she reduced to 1.25%.
Figure 14 Analysis of the consumption in relation to the consumption of reference of the different drivers in both
monitoring phases through the ANOVA1 method
Concerning the consumptions, through the adopted methodology to proceed to a comparison of the
consumptions among the different participants, it is possible to put them in the same graphic display
shown in figure 14. The analysis of the consumption reduction has to do with the drivers change in
previous results, a better explanation will be given in the next chapter.
![Page 12: Monitoring driving behaviour in fuel consumption in light ... · ANOVA1. VSP METHODOLOTY When we deal with real measurements on the road and taking into account the great variability](https://reader034.vdocuments.mx/reader034/viewer/2022052100/6039b51ba27cc523681068f0/html5/thumbnails/12.jpg)
12
Conclusion
Regarding the dynamics the developed models proved to be efficient because through them it was
possible a distinction in driving behaviour of the different participants which enabled us to clearly
distinguish the more aggressive from the more gentle drivers.
Concerning the hypothesis of gentle driving being conditioned by the capacity of the vehicle acceleration,
this idea was rejected. The gentler driver of the group (Driver 4) possesses a vehicle with a less engine
power/weight (0.043W/kg) and he/she is the one who seems to less limited by his/her vehicle not
overtaking the 60% and the 50% of the vehicle capacity in the first and second phase respectively. Driver
4 demanded from his/her vehicle he owns less than half of its response capacity. Driver 1 (apparently the
most aggressive of the group) might be limited, showing situations that demand more than 90% of the
maximum power of the vehicle. This may be read as if the most aggressive driver who possesses the
vehicle with better performance (engine power/weight equal to 0.072) had a vehicle of even more
capacity of acceleration would be even more aggressive. Relatively to others drivers analyzed in
situations at the limit, Driver 2 and Driver 5 attained 70% and Driver 3 reached 80% of the response
capacity of the vehicle. So one concludes that they were not being limited by the acceleration capacity of
their vehicles.
Concerning the analysis about the consumptions, through the adopted methodology to proceed to a
comparison of the consumptions among the different participants, it is possible to put them in the same
graphic display shown in figure 58. We used the statistic method ANOVA1 to evaluate if the behaviours of
the different drivers were statistically different. In the consumption analysis we conclude that all drivers
showed a significantly reduction except Driver 2 and Driver 3. Driver 1 showed the highest reduction
(29,2%), which was expectable since Driver 1 showed the highest improvement in his driving behaviour in
va positive component (25,2%) and va negative component (31,6%) through VSP methodology. In
ANOVA1 methodology the improvement of acceleration and negative acceleration was highest as well,
19,5% and 17,1% respectively. Driver 2, on the other hand, showed hardly any improvement, va positive
component only 3,6% and va negative component 0,8% (VSP analysis) and acceleration rate 2,7% and
negative acceleration rate 4,3%, so in the end result Driver 2 did not show a significantly consumption
reduction (p>0,05). In ANOVA1 analysis for acceleration and negative acceleration rates Driver 3
improved 7,4% and 9,6% and although Driver 3 did show some improvement in va positive component of
11%, he decreased on va negative component with 10,8% in VSP analysis. In the end result it was not
sufficient to reveal a significantly consumption reduction in ANOVA1 method for consumption analysis
(p>0,05). However, it was noticed that Driver 5 also had a decrease in va negative component of 13,5%,
even higher than Driver 3, and Driver 5 showed a significantly consumption reduction of 15,7%. However
Driver 5 revealed higher improvement than Driver 3 in va positive component (19%). And it was revealed
that in ANOVA1 analysis for acceleration and negative acceleration rated Driver 5 improved 15,4% and
12,4% respectively.